﻿################
## Readme for DeScioli and Kimbrough, “Alliance formation in a side-taking experiment”, JEPS
################

Individual session datafiles were compiled to produce the AllSessions_JEPS.csv. The data in AllSessions_JEPS are listed in the order the experiments were conducted.

All analyses are reproduced in two files. The Replication_File_JEPS.R file reproduces most of the analyses in the paper. It also produces a Stata .dta file which is used for the regression analysis. The Replication_Regressions_JEPS.do file reproduces Tables 1 and A1.

The log files show the output from running the .R and .do files. Note that .R files should be opened in the program R, and .do files should be opened in Stata. The .txt files should be opened in a text editor such as TextEdit, Notepad++, WordPad, or MSWord (but the formatting may not be preserved in regular Notepad).

################

List of variables in AllSessions_JEPS.csv
______________________________________
Session = session identifier
Condition = treatment description
ID = redundant session identifier
Fight = not used
Subject = not used
Period = period within session (1st period reports initial randomly assigned ranks, so periods reflecting subjects' choices are 2-21).
TableOn = indicator for Table treatment
ARank1 - ARank8 = Person A's Ranks of Persons A-H
BRank1 - BRank8 = Person B's Ranks of Persons A-H
CRank1 - CRank8 = Person C's Ranks of Persons A-H
DRank1 - DRank8 = Person D's Ranks of Persons A-H
ERank1 - ERank8 = Person E's Ranks of Persons A-H
FRank1 - FRank8 = Person F's Ranks of Persons A-H
GRank1 - GRank8 = Person G's Ranks of Persons A-H
HRank1 - HRank8 = Person H's Ranks of Persons A-H


################

List of variables in Individual_Data_JEPS.csv
______________________________________

id = subject id
session = session id
period = period within session
alliance = correlation between i's ranks of all j at t and j's ranks of i at t-1
egalitarian_bandwagon = correlation between i's ranks of all j at t and sum of others' ranks of all j at t-1
treatment = treatment indicator
chat = chat indicator
full = full info indicator
type = 1 if alliance>0.2, 2 if egalitarian_bandwagon>0.2, 3 if abs(egalitarian_bandwagon)>0.2, else 4

